Joint seminar IPHYS & IEL: Signal processing and Machine Learning in high-dimensions: Computation, Physics and Optics
Event details
| Date | 19.02.2020 |
| Hour | 11:15 › 12:15 |
| Speaker | Florent Krzakala is a professor at Sorbonne Université and a Researcher at Ecole Normale Superieure in Paris. His research interests include Statistical Physics, Machine Learning, Statistics, Signal Processing, Computer Science and Computational Optics. He leads the SPHINX “Statistical PHysics of INformation eXtraction” team in Ecole Normale in Paris, and is the holder of the CFM-ENS Datascience chair and of a PRAIRIE Institute chair. He is also the funder and scientific advisor of the startup Lighton. |
| Location | |
| Category | Conferences - Seminars |
Abstract: The increasing dimensionality of data in the modern computing age presents new challenges and opportunities to the field of signal and data processing. One the one hand, processing these data requires new software and hardware solutions, on the other hand, the high-dimensional settings allow one to use powerful asymptotic methods from probability theory and statistical physics to obtain precise characterizations and develop new algorithmic approaches.
Here I will present some of the results obtained through these methods for two sets of problems: (i) reconstructing a signal from a minimal number of measurements in a variety of situations, including for instance phase retrieval and (ii) analyzing the performances of neural networks and their ability to avoid overfitting. I will also discuss some practical implementations of these ideas for both problems in the context of optics in random media and optical computing.
Here I will present some of the results obtained through these methods for two sets of problems: (i) reconstructing a signal from a minimal number of measurements in a variety of situations, including for instance phase retrieval and (ii) analyzing the performances of neural networks and their ability to avoid overfitting. I will also discuss some practical implementations of these ideas for both problems in the context of optics in random media and optical computing.
Links
Practical information
- General public
- Free
Organizer
- Prof. Jean-Philippe Thiran
Contact
- Prof. Jean-Philippe Thiran, [email protected]/+41 21 693 46 23